Topic : QUERY UNDERSTANDING IN AMAZON SEARCH
Abstract: “Amazon is one of the world’s largest e-commerce sites and Amazon Search powers the majority of Amazon’s sales. A key component of Amazon Search is the query understanding pipeline, which modified appropriate semantic information used to display products for billions of queries Everyday The
Query tagging is the task of semantically annotating query terms to pre-defined labels (such as brand, product-type and color). In this talk, we propose a scalable system to train large-scale machine learning algorithms to solve this problem. A Popular technique employed in NLP literature for tagging parts of text is conditional random fields (CRF). Most work uses supervised approaches to train this model, however, man annotating billions of check seen on Amazon is an extremely expensive task. The training data needs to Be mercurally every, For capable of scaling to billions of queries, where we automatically derive labels for query tokens using additional resources. Specifically, we present a principled approach to obtain derived labels in queries by leveraging user click and purchase data to compute query To product affinities,And structured product catalog data (such as brand of the Product) to obtain query to label affinities. Our system improved the precision over baseline by 10% Our system improved the precision over baseline by 10% Our system improved the precision over baseline by 10%
Bio: Lead Data Scientist, Guided Search at A9.com
Tanvi has been working with Amazon’s search engine technology for the past three years and her primary focus has been understanding users’ queries to improve places of products on Amazon. Prior to That she worked at Adchemy (later acquired by Walmart Labs) and developed their keyword recommendation system for online advertisers. She completed her Masters in Computer Science from The University of Texas at Austin, publishing papers with Prof. Raymond Mooney on problems of Integrating NLP And Computer Vision (NIPS, 2011 and ECAI, 2012). During her masters, she worked at Facebook to improve places of facebook app stories on News Feed. She has been developing scalable, end-to-end,Machine learning systems for search and ads technology for the past 8 years.